SOTAVerified

3D Object Classification

3D Object Classification is the task of predicting the class of a 3D object point cloud. It is a voxel level prediction where each voxel is classified into a category. The popular benchmark for this task is the ModelNet dataset. The models for this task are usually evaluated with the Classification Accuracy metric.

Image: Sedaghat et al

Papers

Showing 51–93 of 93 papers

TitleStatusHype
Regularization Strategy for Point Cloud via Rigidly Mixed SampleCode1
Spherical Transformer: Adapting Spherical Signal to CNNs—0
Generative VoxelNet: Learning Energy-Based Models for 3D Shape Synthesis and Analysis—0
Learning Geometry-Disentangled Representation for Complementary Understanding of 3D Object Point CloudCode1
Classification of Single-View Object Point Clouds—0
I3DOL: Incremental 3D Object Learning without Catastrophic Forgetting—0
A Fast Hybrid Cascade Network for Voxel-based 3D Object ClassificationCode0
Point TransformerCode1
Cascaded Refinement Network for Point Cloud Completion with Self-supervisionCode1
Weight Excitation: Built-in Attention Mechanisms in Convolutional Neural NetworksCode1
Generalized Multi-view Shared Subspace Learning using View Bootstrapping—0
Global-Local Bidirectional Reasoning for Unsupervised Representation Learning of 3D Point CloudsCode1
FPConv: Learning Local Flattening for Point ConvolutionCode1
InSphereNet: a Concise Representation and Classification Method for 3D ObjectCode0
L3DOC: Lifelong 3D Object Classification—0
Data-Free Point Cloud Network for 3D Face Recognition—0
Multi-Task, Multi-Channel, Multi-Input Learning for Mental Illness Detection using Social Media Text—0
Addressing the Sim2Real Gap in Robotic 3D Object Classification—0
Spherical Kernel for Efficient Graph Convolution on 3D Point CloudsCode0
Revisiting Point Cloud Classification: A New Benchmark Dataset and Classification Model on Real-World DataCode1
ClusterNet: Deep Hierarchical Cluster Network With Rigorously Rotation-Invariant Representation for Point Cloud Analysis—0
Spherical Fractal Convolutional Neural Networks for Point Cloud Recognition—0
Octree guided CNN with Spherical Kernels for 3D Point Clouds—0
Extending Adversarial Attacks and Defenses to Deep 3D Point Cloud ClassifiersCode0
3D Point Capsule NetworksCode0
3DTI-Net: Learn Inner Transform Invariant 3D Geometry Features using Dynamic GCN—0
A Graph-CNN for 3D Point Cloud ClassificationCode0
SPNet: Deep 3D Object Classification and Retrieval using Stereographic Projection—0
MeshCNN: A Network with an EdgeCode0
Learning 3D Shapes as Multi-Layered Height-maps using 2D Convolutional NetworksCode0
Spherical Convolutional Neural Network for 3D Point Clouds—0
Deep 2.5D Vehicle Classification with Sparse SfM Depth Prior for Automated Toll Systems—0
General-Purpose Deep Point Cloud Feature ExtractorCode0
3D Object Classification via Spherical Projections—0
O-CNN: Octree-based Convolutional Neural Networks for 3D Shape AnalysisCode1
Wide and deep volumetric residual networks for volumetric image classification—0
Learning a Hierarchical Latent-Variable Model of 3D ShapesCode0
Dynamic Edge-Conditioned Filters in Convolutional Neural Networks on GraphsCode0
ScanNet: Richly-annotated 3D Reconstructions of Indoor ScenesCode1
OctNet: Learning Deep 3D Representations at High ResolutionsCode0
FusionNet: 3D Object Classification Using Multiple Data Representations—0
RotationNet: Joint Object Categorization and Pose Estimation Using Multiviews from Unsupervised ViewpointsCode0
Block Coordinate Descent for Sparse NMFCode1
Show:102550
← PrevPage 2 of 2Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1OursClassification Accuracy93.6—Unverified
2G3DNet-18 MLP, Fine-Tuned, VoteClassification Accuracy91.7—Unverified
3CrossMoCoClassification Accuracy91.49—Unverified
4O-CNN(6)Classification Accuracy89.9—Unverified
5Spherical KernelClassification Accuracy89.3—Unverified
63D-PointCapsNetClassification Accuracy89.3—Unverified
7ECC (12 votes)Classification Accuracy83.2—Unverified
#ModelMetricClaimedVerifiedStatus
1PolyNetAccuracy94.93—Unverified
2ORIONAccuracy93.8—Unverified
3G3DNet-18 SVM, Fine-Tuned, VoteAccuracy93.1—Unverified
4ECC (12 votes)Accuracy90—Unverified
#ModelMetricClaimedVerifiedStatus
1SceneGraphFusionTop-10 Accuracy0.8—Unverified
23DSSG [Wald2020_3dssg]Top-10 Accuracy0.78—Unverified
#ModelMetricClaimedVerifiedStatus
1YOLO-Xmean average precision0.99—Unverified